A major aim of this project is to continue development of our theoretical perspective on speech, according to which phonetic gestures of the vocal tract are at once the fundamental units of the phonology, the smallest units of speech production and the smallest perceivable units of a speech message. A theory of speech perception that is constrained to be compatible with theories of phonology and of speech production, has, we argue, a stronger claim to realism than ones developed unconstrained in these ways. In the first section of the proposal, we propose research on speech production and perception designed to develop Browman and Goldstein's theory of articulatory phonology along two major lines. First, we test a proposal in the theory that gestural, not featural, units serve as units of contrast in languages. Second, we focus on principles by which gestures are relatively timed to form syllables of the language. In a second section of the proposal, we address a major challenge to our view that athe phonetic gestures of articulatory phonology are perceived that takes the form of evidence interpreted as showing that speakers do not attempt to achieve articulatory targets, but rather, acoustic ones. If that is the case, then there is no justification to our proposal that listeners hear gestural targets. In a third section, we propose experiments to test directly our view that gestures, not acoustically-defined phonetic information, is extracted from the acoustic speech signal by listeners. A final section proposes to use behavioral measures and fMRI measures of brain activity to test a claim of the motor theory that gestural perception is achieved by a specialized phonetic module int he human brain.